Raw Strings lack parody

[NicholasWillhite]

BPO	44308
Nosy	@mdickinson, @vstinner, @ezio-melotti, @stevendaprano

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = <Date 2021-06-04.15:33:04.732>
created_at = <Date 2021-06-04.04:21:26.007>
labels = ['type-feature', 'invalid', 'expert-unicode', '3.11']
title = 'Raw Strings lack parody'
updated_at = <Date 2021-06-04.15:33:04.731>
user = 'https://bugs.python.org/NicholasWillhite'

bugs.python.org fields:

activity = <Date 2021-06-04.15:33:04.731>
actor = 'vstinner'
assignee = 'none'
closed = True
closed_date = <Date 2021-06-04.15:33:04.732>
closer = 'vstinner'
components = ['Unicode']
creation = <Date 2021-06-04.04:21:26.007>
creator = 'Nicholas Willhite'
dependencies = []
files = []
hgrepos = []
issue_num = 44308
keywords = []
message_count = 9.0
messages = ['395064', '395066', '395067', '395076', '395077', '395079', '395080', '395086', '395092']
nosy_count = 5.0
nosy_names = ['mark.dickinson', 'vstinner', 'ezio.melotti', 'steven.daprano', 'Nicholas Willhite']
pr_nums = []
priority = 'normal'
resolution = 'not a bug'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'enhancement'
url = 'https://bugs.python.org/issue44308'
versions = ['Python 3.11']

[NicholasWillhite]

I'm really sure this isn't filed correctly. I'm a total noob to this process, so feel free to redirect me. :)

Bytes can be defined as a function, or a prefixed series. You can prefix a series with "b" and get the expected data type. You can also use the builtin functions "bytes" to get the same structure:

bytes('foo\bar', 'utf-8') == b'foo\bar'
True

But there's no builtin function for r'foo\bar' that gives you 'foo\\bar'.

This would be really handy for applications that accept a regular expression. If that regex was part of the source code, I'd just r'foo\bar' to get the expected string. Being able to accept something like bytes and do:

  data = b'foo\bar'
  raw_string(data)
  'foo\\bar'

would be really useful for applications that accept a regex as input.

Is there an obvious way to do this that I'm not seeing? Has my google-foo failed me? Feels like a function that should exist in the stdlib.

Again, really sure I'm not "doing this correctly." So please direct me! :)

Appreciative,
-Nick Willhite

[stevendaprano]

I think you have missed something important here:

    >>> data = b'foo\bar'
    >>> len(data)
    6
    >>> print(data)
    b'foo\x08ar'

If you want bytes including a backslash followed by a b, you need to use raw bytes rb'foo\bar' or escape the backslash.

Also Python 3.8 is in feature-freeze so the earliest this new feature could be added to the language is now 3.11.